Tracking dead clients..

Hi,
A newbie question. How can a server figure out if a client has been improperly shutdown/has crashed/is dead ? Is there a way of pinging the client ?
thanks in advance.
Ganesh

This is a good approach, but there's a slight problem: the RMI Runtime will only call this method when the list of client objects AS A WHOLE no longer references the remote object! If you wanted information about only a CERTAIN client, then this approach does not work, since the runtime cares only if ALL of the apparent objects that could or/are refencing a certain bound object, are NOT referencing that object TOGETEHER. Moreover, Sun does not specify after how long the "unreferenced" method will be called. This is another disadvantage. If you really must go ahead with your little project, then i suggest you build in your own technique to detect both individual client crashes/non-references. If you do, you even have the advantage of building a mechanism whose time can be altered (that is, the mechanism to determine AFTER HOW LONG the unreferencing should be checked.)

Similar Messages

How to keep track of client in rmi server

how can I keep track of a disconnected client in a rmi server ?
And what logic do I use to disconnect a client fro mthe server ?

When a RMI client receives a stub to the remote object, RMI client runtime sends a dirty() call to the RMI server runtime which in turn helps the RMI server runtime in updating the client references to the remote object. Client reference is nothing but a unique identification generated by client RMI runtime for that instance of JVM. RMI uses a reference counting algorithm for destroying unreferenced remote objects. Once all references (all clients who are holding the stub to this remote object) to a remote object drops to zero, local garbage collector takes care of freeing this remote object.
RMI client runtime initiates a TCP connection with the remote object only when a method is invoked on the stub. RMI transport layer uses a simple connection pooling mechanism to use already established TCP connections with server objects in further remote method invocations. Let's say the client is invoking a remote method on the remote object's stub fetched from the registry service. Now RMI transport layer initiates a TCP connection with the remote object (exported at ("host1", 4567)). After completion of the remote method, RMI client's transport layer keeps the connection open for some time (some configurable value) to take the advantage of the already established TCP connection in future remote method invocations on remote objects exported on the same host and port ("Host1", 4567"). RMI transport layer also takes care of closing these connections if the connections are idle more than configured value of connection open time.
From the RMI server's perspective, once the RMI server runtime receives a remote method invocation request from the RMI client, it creates a new thread and try to dispatch the method on the designated remote object. After returning the results of the remote method to the client, RMI client's transport layer may not end this conservation with this remote object to take the advantage of this TCP connection in future remote method invocations. So the same thread may be used for dispatching another remote method in future. This is actually the side effect of re-using the established TCP connections on the client side.
RMI server maintains client's reference in the client's reference set of the remote object as long as the client holds a normal reference to remote object's stub in its JVM. Once the client relinquishes the reference to the stub (goes out of the scope or setting it to the null), RMI client automatically sends clean() call to the RMI server runtime to update client reference set of the remote object. There is no public API to lookup or manipulate the client references for a given remote object, but you can always unexport the remote object even some clients are holding stubs to the remote object. Next time when the client invokes the remote method using this stub, client will receive a "connection refused exception" from the server because the remote object is already unexported. Let's say if the RMI server is started again (remote object is exported and registered with registry service), the client once again obtains the stub to the remote object from the registry service instead of using earlier stale stub. (This time remote object may have been exported on different port, so the port details in the old stub may not be valid any more).
At any point of time there can be more than one established TCP connection with the RMI client. This may be the temporary condition; RMI client's transport layer automatically closes these connections if it does not receive any remote method invocation requests with in a pre-determined interval of time (default: 2 min). You can always find out RMI client's host in a given remote invocation by using getClientHost(), but how can you semantically define a "disconnected client in a rmi server" in your application context ? You can always force all the clients to fetch the stub again in future remote method invocations by unexporting the remote object in RMI server. I hope this information helps ...
-- Srinath Mandalapu

Checking on dead client

Hello all,
I have written a custom datasource to stream jpeg images usinf RTP to an applet in a web page.
My question is: How do I check to see if my client is still alive? If I just close the browser, my application has no way of knowing that it should stop. I am using an RTPManager and thought that I would have something in that class that would signal this condition, but haven't run across it yet.
Thanks in advance,
Bill Bedsaul

you can overide the finalize method of the applet like
this:
protected void finalize() throws Throwable {
//put your code to close the streams and resources
here
this method should get called when closing the browserCan this method be used if suddenly there is a problem with the network?

Need software to keep track of clients for small business.

We need some way to keep up with multiple types of payments for several hundred clients, as well as the ability to make notes as needed. We have been using ical for our notes and a simple Numbers spreadsheet, but this method is cumbersome. I have looked at various accounting programs, but they all seem much more complicated than I need. Any help or suggestions are greatly appreciated!

I ran a small business, and there is no way I would have done it with iCal and Numbers. Intuit's Quickbooks is the best program for small business. Yes, it's probably got more than you need, but it helps pay bills, taxes, inventory, blah blah blah. It's only $200. They have an online version. It links to your bank. It links to tax software. I can go on. I used the enterprise version, and I'm not an accounting, and I found it a piece of cake to learn.
Since this is the forum for the Macbook Pro, you're really not going to find much info here. Go to the Intuit Quickbooks website, and do a test run. Trust me, if you want to make a profit and pay bills on time, get a real program. Quickbooks is one of the cheapest but still the best out there.

Alert server down

Hi friends,
We are facing an issue alert server is down so in sapgui we could see the alert server is in red.Also one of my colleagues told we need to call the TREX environment before starting the TREX.
When i tried to start the alert server i am getting an error message as
empty PIDs method RESTART/KILL not supported
Return code 4910 )
Kindly help me in this.How to call TREX environment or any other solution????????
Thanks,
Naga Sankar Anand

Yes. It is possible. Look at the "Tracking Dead Clients" post just a few posts below yours. We were going back and forth this other guy that had the same problem as you.

Keeping track of current clients on server

My problem is that i have all these clients connected to my server. I don't want the same client connecting to the server if he is already connected.
I was thinking that I could keep track of who is currently connected on the server side, but that leaves many gaps. What happens when a client crashes, and he just drops off the server. The server won't know this happened, and it would be hard to know which client dropped off.
My next idea was that the server could send a query out to the clients, and each client that threw a IOException would be a client that wasn't logged on anymore. and i could cut his server connections, and end his thread.
Is there maybe a better way to keep track of clients currently connected to the server?

The idea of probing the clients can be a bad one, because it takes some time for the server to figure out that a client isn't on. The query you send out will be the same as any other data, and if the query could find out the connection is down, then the connection would already have been broken somewhere else, and you'd know the user is offline. (Assuming you have sane coding practises, which is an IOException handler around the main data loop, and a finally{} block which marks the user as disconnected.)
One way to ensure that the same user isn't connected twice is to boot off the first instance when the second instance connects. Regrettably a lot of company's solutions to this problem are the other way around, and block you from logging into the server until the first connection times out (a good example of this is the MMOCRPG, Ragnarok Online, where getting disconnected means having to wait about 60 seconds until the server realises your first connection is gone. ;-))

Cannot see clients on maps

Hello,
We have location servers running 5.2.91.0 and WLCS running 5.2.178 but cannot track any clients-see attached .location server is contactable from WCS & WLCs ,NTP is configured,LOCP,SNMP,SOAP all permitted on the firewall between the controllers & servers (WCS & location are on the same subnet ). Checked IPtables on WCS and its traffic is permitted there as well.It was working fine when we had 4.2 code on WLCs.
Any help is much appreciated.
Thanks,
janesha

What version is WCS?
When you upgraded WLC code, you need to resync LOC server on WCS so it pushes the Self signed cert between location server and WLC.
Do the 2 following things:
SSH to location server and run:
/opt/locserver/bin/getserverinfo -> see if it displays values for tracked clients.
If it does and you still dont see them, delete and re add the location server from WCS, that should fix your issue.

IP device tracking

Hi,
We have Cisco 3850 switches and we dont use dot1x but we need to turn off ip device tracking but when I do it from global config mode it pops up the below error:
Switch(config)#no ip device tracking
% IP device tracking is disabled at the interface level by removing the relevant configs
I've tried disabling it under interface mode even though we don use it.
Can someone please show me how to disable it globally?
Thanks.

I have tried the no ip device track max 10 in interface mode and it accepts it but when I issue " sh ip device tracking int gig 2/0/22 " it still says its enabled.
SW#show ip device tracking interface gig 2/0/22
Enabled interface Configs:
Global IP Device Tracking for clients = Enabled
Global IP Device Tracking Probe Count = 3
Global IP Device Tracking Probe Interval = 30
Global IP Device Tracking Probe Delay Interval = 10
IP Address    MAC Address   Vlan Interface           Probe-Timeout      State    Source
Total number interfaces enabled: 64
Enabled interfaces:
Gi1/0/1, Gi1/0/2, Gi1/0/3, Gi1/0/4, Gi1/0/5, Gi1/0/6, Gi1/0/7,
Gi1/0/8, Gi1/0/9, Gi1/0/10, Gi1/0/11, Gi1/0/12, Gi1/0/13, Gi1/0/14,
Gi1/0/15, Gi1/0/16, Gi1/0/17, Gi1/0/18, Gi1/0/19, Gi1/0/20, Gi1/0/21,
Gi1/0/22, Gi1/0/23, Gi1/0/24, Gi1/1/1, Gi1/1/2, Gi1/1/3, Gi1/1/4,
Te1/1/1, Te1/1/2, Te1/1/3, Te1/1/4, Gi2/0/1, Gi2/0/2, Gi2/0/3,
Gi2/0/4, Gi2/0/5, Gi2/0/6, Gi2/0/7, Gi2/0/8, Gi2/0/9, Gi2/0/10,
Gi2/0/11, Gi2/0/12, Gi2/0/13, Gi2/0/14, Gi2/0/15, Gi2/0/16, Gi2/0/17,
Gi2/0/18, Gi2/0/19, Gi2/0/20, Gi2/0/21, Gi2/0/22, Gi2/0/23, Gi2/0/24,
Gi2/1/1, Gi2/1/2, Gi2/1/3, Gi2/1/4, Te2/1/1, Te2/1/2, Te2/1/3,
Te2/1/4
Here is the show version:
SW#show ver
Cisco IOS Software, IOS-XE Software, Catalyst L3 Switch Software (CAT3K_CAA-UNIVERSALK9-M), Version 03.02.01.SE RELEASE SOFTWARE (fc1)
Technical Support: http://www.cisco.com/techsupport
Copyright (c) 1986-2013 by Cisco Systems, Inc.
Compiled Wed 20-Mar-13 17:10 by prod_rel_team
Cisco IOS-XE software, Copyright (c) 2005-2013 by cisco Systems, Inc.
All rights reserved. Certain components of Cisco IOS-XE software are
licensed under the GNU General Public License ("GPL") Version 2.0. The
software code licensed under GPL Version 2.0 is free software that comes
with ABSOLUTELY NO WARRANTY. You can redistribute and/or modify such
GPL code under the terms of GPL Version 2.0.
(http://www.gnu.org/licenses/gpl-2.0.html) For more details, see the
documentation or "License Notice" file accompanying the IOS-XE software,
or the applicable URL provided on the flyer accompanying the IOS-XE
software.
ROM: IOS-XE ROMMON
BOOTLDR: C3850 Boot Loader (C3850-HBOOT-M) Version 1.1, RELEASE SOFTWARE (P)
SW uptime is 4 weeks, 1 day, 19 hours, 3 minutes
Uptime for this control processor is 4 weeks, 1 day, 19 hours, 6 minutes
System returned to ROM by reload at 12:43:29 WST Sun Sep 8 2013
System restarted at 13:08:55 WST Sun Sep 8 2013
System image file is "flash:packages.conf"
Last reload reason: Reload command
This product contains cryptographic features and is subject to United
States and local country laws governing import, export, transfer and
use. Delivery of Cisco cryptographic products does not imply
third-party authority to import, export, distribute or use encryption.
Importers, exporters, distributors and users are responsible for
compliance with U.S. and local country laws. By using this product you
agree to comply with applicable laws and regulations. If you are unable
to comply with U.S. and local laws, return this product immediately.
A summary of U.S. laws governing Cisco cryptographic products may be found at:
http://www.cisco.com/wwl/export/crypto/tool/stqrg.html
If you require further assistance please contact us by sending email to
[email protected].
License Level: Ipbase
License Type: Permanent
Next reload license Level: Ipbase
cisco WS-C3850-24P (MIPS) processor with 4194304K bytes of physical memory.
Processor board ID FOC1722Z4J9
2 Virtual Ethernet interfaces
56 Gigabit Ethernet interfaces
8 Ten Gigabit Ethernet interfaces
2048K bytes of non-volatile configuration memory.
4194304K bytes of physical memory.
250456K bytes of Crash Files at crashinfo:.
250456K bytes of Crash Files at crashinfo-2:.
1609272K bytes of Flash at flash:.
1609272K bytes of Flash at flash-2:.
0K bytes of Dummy USB Flash at usbflash0:.
0K bytes of Dummy USB Flash at usbflash0-2:.
0K bytes of at webui:.
Base Ethernet MAC Address          : d0:c7:89:70:a7:00
Motherboard Assembly Number        : 73-12240-10
Motherboard Serial Number          : FOC17215VEG
Model Revision Number              : B0
Motherboard Revision Number        : D0
Model Number                       : WS-C3850-24P
System Serial Number               : FOC1722Z4J9
Switch Ports Model              SW Version        SW Image              Mode
     1 32    WS-C3850-24P       03.02.01.SE       cat3k_caa-universalk9 INSTALL
     2 32    WS-C3850-24P       03.02.01.SE       cat3k_caa-universalk9 INSTALL
Switch 02
Switch uptime                      : 4 weeks, 1 day, 19 hours, 6 minutes
Base Ethernet MAC Address          : d0:c7:89:70:96:80
Motherboard Assembly Number        : 73-12240-10
Motherboard Serial Number          : FOC17215V33
Model Revision Number              : B0
Motherboard Revision Number        : D0
Model Number                       : WS-C3850-24P
System Serial Number               : FOC1722V19Q
Configuration register is 0x102

Tracking logon user's IP address

Hi gurus,
We are implementing an external facing portal in EP 7.0. As part of audit trail, we need to track the IP address of the user logging into the portal. While I have been going through the default activity report provided + configuration of new reports, nowhere do i find the mention of tracking the client IP address. I went thru Prakash's blog on Writing your own portal statistics tool, and was wondering if we do have to take up such a customization in logo.par?
I am seeking suggestions on how to track and record the client IP address once a user logons into the portal.
regards,
S Rudra

Hi,
Thanks for the single reply to the question .. not to mention awarded good points too ... configuring the way mentioned gives the details into the standard log file.... one has top again filter out the details
I solved the problem myself ... HOW .... here it goes
But I was looking to get custom report, to get the login userid, IPaddress and time of access.... built a custom class to write the details into a text file and stored it into a server location ... then I modified the masthead par to implement the same ... not the logon.par ... currently working to get the output in download excel file ...
Cheers!!!!

Difference between delete aged discovery data and delete inactive client discovery data maintenance tasks

I'm setting up a new CM12 site and one issue with our current CM2007 site that I'm trying to minimise in this new site is the discovery and retention of dead clients.
I think the new options as part of the system and group discoveries to only discover systems that have logged on to a domain/updated their password in a set period will resolve the majority of the problems we have here. I've set both of these to 30 days.
Management are keen to be pretty aggressive with the retention period so we're looking to only keep computers in the database that have heartbeated within 30 days. I've set the heartbeat discovery to 8 hours and now I'm looking at the maintenance tasks.
The task I think I need to configure is 'delete inactive client discovery data'. I've set this to 30 days and to run every Saturday. If I look at the 'delete aged discovery data' task it's set to 90 days every Saturday. How will that affect things?
Do I need to make both tasks match?
P.S. Thoughts on the retention period...is 30 days too aggressive? My original design was 60 days.

I do have question about this and more related to "Delete Aged Discovery Data". I've got following configured in my environment for AD Discovery (Development environment)
do not discover machine not logged on for 80 days | do not discover machines not changed password for 80 days.
if I run the query select * from v_r_system (nolock) where datediff("dd", last_logon_date0, getdate()) > 80
it returns about 600 devices.
The task "Delete Aged Discovery Data" is also configured to remove anything that has not been discovered in last 80 days. When this task run I would expect that the 600 devices should be removed from the db, however it only removed 1 device ???
question I have
do you know which other criteria is taken into consideration for this to work. I'm currently using SQL Profiler to determine which StoredProcedure/Function is used to remove these devices.
this is my development environment I only have 2-3 clients all the other devices are just discovered with no SCCM Client installed.
do the devices have to be discovered as part of AD Device discovery or is it sufficient that they get discovered through AD Group membership, in our environment AD Device discovery will not work as we don't have WINS/ DDNS enabled.
thx for any help.

How to remove old clients in MSE3350

We have reached the limit off 3000 clients in our MSE 3350. This is moslty do to a lot of probing clients, BUT we need to track probing clients.
How ever we believe that this large number must include clients that is no longer active.
How can we change how fast a clietns that is no longer seen is removed from the MSE leaving space to new clients.
Best regards Rasmus Sindum, NetDesign

Hi,
on the WLC, the probing client is deleted after 5 minutes. So check out the number of clients on your WCS and if it is 3000, then it means you just have too many people "just probing".
Regards,
Nicolas

When client crashes...

Hi
I've a server program and a client program (RMI). In the client I've a method to disconnect from the server using server.unregister(this) and UnicastRemoteObject.unexportObject(this, true). This works fine, but when I eliminate the client using Windows TaskManager, I can't start the client program again on the same PC.
How do I solve this problem? Is there a way to let the server release the dead client(s)?
Regards,
Franck

well if you kill the client via tskmgr then of course the Server will think it still alive, cos it wont have done any cleanup....
the only thing you can do it build a "ping" routine into the client and server...... if a client doesn't ping the server within your time period then automatically unregister and export the client.......
it's simple keep alive routine most network programs use......

Customer tracking problem

Hi guys,
I am doing a Yoga Studio Website. For the most part it is going pretty straight forward, but the issue is they need a customer tracking feature built in that they can just scan a clients card and it records that they were there and and reduces their number of pre purchased classes.
I can do this very easily with PHP/MySQL. However I see no way to do this with Business Catalyst.
So is there a way that I can pull the customer contact data from the BC site into an external php site that I will have to build for them to track their client visits?
Having to enter client info twice into two different systems is not a good option in this case.
Any ideas?

Hey BRad.
Use the BC data but API.
make a custom field which is their pre-purchase classes. They can pay and make request for them on the site and admin can update the field there.
API wise you external system pics up the scanning links directly to you middle layer (PHP) which processes and updates BC for them.
The Data needs a unique identifier so their email or I'd so you can look for them on the BC site and then update the data.
Or it could be done manually in the CRM when ever that happens.

Move to start of next Track?

Am building a corporate DVD that has MainMenu > 10 submenus > 10 tracks.
Client wants to use the DVD remote advance button >>| to move from each item (slides & video) to the next within a track, while each item 'holds' until advance button >>| is pressed - no problem, just add a marker each time and set marker playback option to Infinite.
But when moving from last item on a track to first item of next track.... problems! (Client wants ideally to be independent of returns to submenus)
Have tried
a. setting the last item's marker to end jump to the next item's marker
b. setting the last item's marker to end jump to next item's submenu
c. leaving it Not Set
hoping advance button >>| would take us there. But am getting random results - route 'a' above works in half of cases.
Running close to deadline - any help appreciated!

It seems to me that you could put the contents of your 10 tracks onto a single track and then create 10 stories, one story to replace each of your 10 tracks. The first story would include all the content of the track, the second story would contain the content that had been on tracks 2 through 10, the third story would contain the content that had been on tracks 3 through 10, etc.

Prime Infrastructure 2.2 - Wired Clients and Trunk Ports

We have our VMWare ESX hosts connected to our server access switches via trunk ports. Prime doesn't seem to track clients on trunk ports (to avoid showing clients on uplink ports between switches, I'm sure). Since these are not switch-to-switch connections, is there a way to enable Prime to track wired clients on these specific trunk ports so we have MAC/IP client info in Prime for our virtual environment?

Hi,
PI discards all the MAC table entries that are on trunk ports for a switch. This enhancement was added from PI 2.1 & later.
- Ashok
Please rate the useful post or mark as correct answer as it will help others looking for similar information

Tracking dead clients..

Similar Messages

Maybe you are looking for